Challenges of non-commercial multicentre North-South collaborative clinical trials

Authors


Corresponding Author Raffaella M. Ravinetto, Institute for Tropical Medicine, Nationalestraat 155, 2000 Antwerpen, Belgium and Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Belgium. E-mail: rravinetto@itg.be

Abstract

The last decade has witnessed a substantial increase of multi-centre, public health-oriented clinical trials in poor countries. However, non-commercial research groups have less staff and financial resources than traditional commercial sponsors, so the trial teams have to be creative to comply with Good Clinical Practices (GCP) requirements. According to the recent experience of a large multicentre trial on antimalarials, major challenges result from the complexity of multiple ethical review, the costs of in-depth monitoring at several sites, setting up an adequate Good Clinical Laboratory Practices (GCLP) framework, lack of insurers in host countries, and lack of adequate non-commercial data management software. Public research funding agencies need to consider these challenges in their funding policies. They also could support common spaces where North-South collaborative research groups may share critical information, such as on research insurance and open-source, GCP-compliant software. WHO should update its GCP guidelines, which date back to 1995, to incorporate the perspectives and needs of non-commercial clinical research.

Abstract

Objectif

Explorer les moyens de lutter contre Chrysomya putoria, la mouche africaine des latrines, dans les latrines à fosse. Comme les latrines sont une source importante de production pour ces mouches, l’élimination de ces sites de reproduction importants est susceptible de réduire les populations de mouches du village et peut réduire la propagation des agents pathogènes diarrhéiques.

Méthodes

Nous avons traité 24 latrines dans un village gambien: six avec soit (1) du pyriproxyfène, une formulation imitant une hormone juvénile des insectes appelée Sumilarv® 0.5G, à granules de 0,5% de pyriproxyfène, (2) des billes de polystyrène gonflées (EPB), (3) du savon local ou (4) aucun traitement i.e. témoins. Les mouches ont été recueillies à l'aide de pièges de sortie placés au dessus des orifices des fosses, chaque semaine pendant cinq semaines. Dans une autre étude, nous avons testé si les latrines pouvaient servir également comme pièges à mouches efficaces en utilisant les odeurs fécales comme attractifs. Nous avons construit six latrines à fosse chacune avec un piège à mouche intégré et avons testé leur efficacité de capture par rapport à six pièges appâtés sous forme de boîtes de poisson placées à 10 m de la latrine. Des discussions de groupe focalisées ont été menées par la suite pour évaluer l'acceptabilité des latrines attrape-mouche.

Résultats

Le nombre de C. putoria émergents a été réduit de 96,0% (IC95%: 94,5 à 97,2%) 4 à 5 semaines après le traitement avec le pyriproxyfène, de 64,2% (IC95%: 51,8 à 73,5%) après traitement avec du savon local, de 41,3% (IC95%: 24,0 à 54,7%) 3 à 5 semaines après traitement avec des EPB. Les pièges à mouches placés sur latrines ont collecté C. putoria et étaient jugés acceptables par les collectivités locales.

Conclusions

Sumilarv 0,5 G est prometteur comme agent de lutte chimique, tandis que les pièges appâtés aux odeurs de latrines peuvent s'avérer une méthode utile non chimique pour lutter contre les mouches. Les deux méthodes justifient un développement supplémentaire afin de pouvoir réduire la production de mouches à partir des latrines à fosse. Une combinaison des interventions peut s'avérer efficace pour le contrôle des mouches des latrines et des maladies qu'elles transmettent.

Abstract

Objetivo

Explorar las formas de controlar Chrysomya putoria, la mosca Africana de letrinas, en letrinas de hoyo seco. Puesto que las letrinas de hoyo seco son una fuente importante de estas moscas, eliminar estos lugares de reproducción puede reducir las poblaciones de la mosca y por lo tanto reducir la diseminación de patógenos diarreicos.

Métodos

Hemos tratado 24 letrinas en un poblado de Gambia: cada grupo de seis se trató con una de las siguientes (1) piriproxifen, que mimetiza una hormona juvenil del insecto formulada como Sumilarv® 0.5G, y piriproxifen granulado al 0.5%, (2) pastillas de poliestireno expandido (PPE), (3) jabón local (4) sin tratamiento como controles. Las moscas se recolectaron utilizando trampas de salida localizadas sobre los orificios de entrada, semanalmente, durante cinco semanas. En un estudio separado evaluamos si las letrinas también funcionaban como trampas eficientes utilizando los olores fecales como atrayentes. Construimos seis letrinas de fosa, cada una con una trampa para moscas incluida, y evaluamos su eficiencia para atrapar las moscas en comparación con seis cajas-trampa con pescado como cebo, localizadas a 10 m de la letrina. Se llevaron a cabo grupos de discusión focalizada después para evaluar la aceptación de las letrinas con trampas para moscas.

Resultados

El número de C. putoria emergentes se redujo en un 96.0% (IC95%: 94.5–97.2%) 4–5 semanas después del tratamiento con piriproxifen; en un 64.2% (IC 95%: 51.8–73.5%) después del tratamiento con jabón local; en un 41.3% (IC 95% = 24.0–54.7%) tras el tratamiento con PPE, 3–5 semanas después del tratamiento. Las trampas de moscas localizadas en las letrinas recolectaron C. putoria y fueron consideradas como aceptables por las comunidades locales.

Conclusiones

El Sumilarv 0.5G parece ser prometedor como agente de control químico, mientras que las letrinas con trampas de olor podrían ser útiles como método no químico para el control de las moscas. Ambos métodos garantizan un mayor desarrollo para reducir la producción de moscas de las letrinas de fosas. Una combinación de intervenciones podría ser efectiva para el control de las moscas de letrinas y la enfermedad que transmiten.

Introduction

The last decade has witnessed a substantial increase in the number of clinical studies in poor countries. These include commercial trials for marketing authorisation applications in other countries/continents (Department of Health & Human Services 2010; European Medicines Agency 2012) and non-commercial trials addressing relevant public health questions. In particular, the number of multicentre studies by independent North-South research consortia has increased, also thanks to new and innovative product development partnerships and funding mechanisms. Here, we illustrate the challenges faced by these new actors from the experience of a trial we have recently completed.

The 4-ABC trial (The Four Artemisinin-Based Combinations (4ABC) Study Group 2011) was a head-to-head comparison of four antimalarial treatments conducted at 12 sites in seven sub-Saharan countries (Burkina Faso, Gabon, Mozambique, Nigeria, Rwanda, Uganda and Zambia) (Table 1). The trial was funded by the European and Developing Countries Clinical Trials Partnership (EDCTP). Overall, 4116 children with uncomplicated P. falciparum malaria were recruited and followed up for 7 months. The trial was steered by a Co-ordination Committee of representatives of all partners, but the day-to-day scientific coordination was delegated to the coordinating investigator and the field coordinator. The Institute of Tropical Medicine Antwerp (ITM) took responsibility for the legal sponsorship and hosted the Trial Management Group (TMG), which brought together all the ‘central’ operational staff: scientific coordinators, project manager, data manager, clinical trial assistant, financial administrator, laboratory coordinator and study monitor. The TMG ensured the coordinated execution of all activities and the documentation of key decisions, including integrating the inputs from the study sites. The preparatory phase began in December 2005 when the contract with EDCTP was signed; recruitment started in July 2007 and the clinical follow-up was completed by mid-2009. The database was locked in June 2010.

Table 1. Study treatment to be tested by country
CountrySitesStudy treatments
Burkina FasoNanoroASAQDHAPQAL
GabonFougamou, LambarenéASAQDHAPQAL
NigeriaAfokang, PamolASAQDHAPQAL
ZambiaNdolaASAQDHAPQAL
RwandaRukaraDHAPQCD + AAL
RwandaMasheshaDHAPQCD + AAL
UgandaJinjaDHAPQCD + AAL
UgandaTororoDHAPQCD + AAL
MozambiqueManhiçaASAQCD + ADHAPQ
UgandaMbararaASAQCD + ADHAPQ

Organisational and contractual aspects

The long duration of the pre-study period (December 2005–July 2007) was due to the complexity of preparatory activities, ranging from the finalisation of the organisational and contractual aspects to the full set-up of the study sites and procedures. Although such organisational and contractual aspects are common to any multicountry clinical research programme, they may present a major challenge for an academic sponsor with a small staff, particularly when compared with most commercial sponsors. For instance, when the preparation for the 4ABC trial started, the Clinical Trials Unit at the ITM comprised the equivalent of 3.5 full time staff, that is, a project manager, a data manager, a biostatistician and an administrative assistant, who were also dealing with other ongoing projects. In commercial multicentre trials, the range of activities is spread across different specialised units such as planning, regulatory affairs, monitoring, database development and validation, data review and data cleaning, pharmacovigilance. In non-commercial multicentre trials with an academic sponsor, these are often concentrated in a small unit and sometimes on a single person.

The procurement of investigational medicinal products (IMPs) is a good example of additional difficulties met by non-commercial sponsors. While in commercial trials, the sponsor and the manufacturer of the IMP are often the same organisation, in non-commercial trials, the sponsor is not the owner of the products, which must either be bought or be provided by their respective manufacturers. In the case of the 4ABC trial, the IMPs came from four manufacturers, requiring individual contractual and procurement agreements and parallel arrangements for the shipment to the sites, which increased the administrative workload and often caused unplanned delays. Notably, a double blind study design had to be ruled out because of the difficulties and the costs of a double-dummy – a relatively common problem for non-commercial studies (Christensen & Knop 2012), for which it is difficult to foresee concrete solutions.

Multidisciplinary expertise, efficient prioritization and communication among the study partners and the sites are therefore required to fill in the gap and to comply with all good clinical practices (GCP) requirements. Noteworthy, these requirements were defined in the early 1990s (World Health Organization 1995; International Conference of Harmonization 1996) in relation to the capacities of the traditional commercial sponsors and to date they have not been updated.

Study insurance

The sponsor provided a no-fault insurance policy to cover any harm caused by participation in the trial. The lack of public guidance on trials' insurance made it difficult to negotiate the contract, for example, aspects related to the maximum amount per patient and total compensation. Due to the difficulty to find insurance companies in the study countries, insurance was contracted in Belgium (the sponsor's country) rather than in sub-Saharan Africa, where patients were recruited, treated and followed up. As no compensation for trial-related harm was claimed during the study, it remains unclear (and should be further investigated) whether an insurance policy stipulated by a company located outside Africa, with significant language and legal differences to the study countries, may work efficiently.

As proper compensation mechanisms are essential to fulfil the ethical obligation of protecting patients, particularly in vulnerable populations, we also suggest that noncommercial consortia should create spaces to share critical information, such as contract templates, examples of how fees are calculated and a database of experienced insurance companies in low-income countries to provide study specific insurance or ‘umbrella insurances’.

Multiple ethical reviews

The length of the ethical review process was particularly unpredictable. The initial clinical protocol and the subsequent amendments were sequentially submitted to the ITM Institutional Review Board (IRB), to the competent Ethics Committee (EC) in the country of the sponsor and to the IRBs/ECs and competent authorities (CA) in the study countries, a total of 20 bodies. Multiple ethical review was undoubtedly beneficial because of the clear North-South complementarities. Ethical aspects related to indemnification for harm, insurance and confidentiality were highlighted by the Northern ECs, while the comments of ECs in the South focused on the need to ensure the co-ownership of the study data, the study sites' qualification/capacity, the transfer of biological samples abroad and the appropriateness of patients' travel reimbursement (Ravinetto et al. 2011). But, the large number of bodies, the multiplicity of procedural requirements (including different policies on the ethical fees) and the unpredictable timelines for some committees caused delays in achieving the expected milestones.

For North-South collaborative research, we believe that a common process is needed in which different ECs reviewing the same protocol communicate, build on common practices and jointly address conflicting opinions. Such harmonisation of procedures would be beneficial for any other commercial and non-commercial multicentre study.

Monitoring

Monitoring the study conduct, including laboratory activities, presented significant challenges. It soon became evident that the external budget provided for monitoring was not sufficient to meet expectations, particularly during the early months of recruitment, when frequent visits can detect systematic mistakes and anticipate structural problems in a timely manner. Similarly, early visits from the laboratory coordinator are critical to harmonise quality assurance in multicentre trials and to build a comprehensive good clinical laboratory practices (GCLP) (WHO Good Clinical Laboratory Practices 2009) framework for laboratories with less research experience. Unfortunately, no specific budget was available for these activities.

Research groups and donors need to appreciate the relevance of early and timely monitoring of the clinical and laboratory aspects of multicentre trials. They should plan adequate human and financial resources to meet these tasks, based on the complexity and inherent risk of each study, on the relevance of laboratory results for the efficacy and safety outcomes and on the sites' specific context. Research consortia could, in parallel, set up alternative or complementary measures, allowing the verification of the completeness, accuracy and coherence of the study data, even when resources for external monitoring are very small. This may include mechanisms for internal quality control and for reciprocal monitoring schemes. In the first case (often called ‘internal monitoring’), data entered in the case report form by an investigator are formally double checked against source documents by a second investigator or a study nurse. This system has the advantage of working continuously, allowing timely detection of problems and mistakes, though it should not substantially raise the workload of the study staff, as this would in turn lower the quality of the data collected. In reciprocal monitoring schemes, institutions in a research partnership could agree on common monitoring standards and procedures. Interested qualified staff could be offered training in clinical and laboratory monitoring, and exchange monitoring visits between different sites could take place (Chilengi et al. 2010). Such a system would be more expensive than internal monitoring, but still cheaper than external monitoring by commercial contract research organisations (CROs); in addition, it would allow mutual learning. The two systems are not mutually exclusive and could be used in parallel.

Data management

The main challenges during clinical follow-up until database lock were related to collecting and managing the trial data. Given the size of the database (overall 4 000 000 data points), data entry was performed at the sites by ad hoc trained clerks, and data were then transferred to the ITM server. As most sites lacked a sufficiently stable Internet connection, commercial software (Macro®; Infermed) was used as it offered the possibility of entering off-line the data in the electronic case report form (eCRF). This set-up worked satisfactorily, and its efficiency improved when some training and data review activities could be delegated to skilled data managers in the South, as in Burkina Faso (task decentralization allows earlier queries and timely data cleaning).

Therefore, we strongly encourage North-South collaboration in clinical data management (van Loen et al. 2011). However, depending on the commercial software, full capacity transfer to the South cannot be achieved; an open-source GCP-compliant software that can work off-line is urgently needed.

Perspectives for the future

In conclusion, a multicentre clinical trial is a challenging undertaking, particularly for independent research groups, which have fewer human and infrastructural resources than commercial sponsors and have to concentrate a variety of specialised tasks on a handful of people. The low flexibility of the funding obtained from public donors for this purpose makes the task even more difficult. Nevertheless, it is important that non-commercial trials are carried out in compliance with the appropriate ethical principles (World Medical Association Declaration of Helsinki 2008; Nuremberg Code 1947; CIOMS 2002; The Belmont Report, 1979) and methodological standards, to ensure protection of trial subjects and their communities and to guarantee the quality of data and results. Different actors may facilitate this process, firstly by ensuring that sufficient resources are mobilised and secondly by allowing the adoption of appropriate, cost-effective quality assurance tools. In particular:

  • Public research agencies should be sensitised about the inherent difficulties of non-commercial trials, and their policy of granting funds adapted accordingly.
  • These agencies could support or facilitate the development of common spaces where North-South collaborative research groups can share key information such as on insurance and on open-source GCP-compliant software.
  • Non-commercial research institutions and sponsors could develop mechanisms to improve the long-term efficacy and quality of independent clinical research, for example, facilitating the dialogue among ECs in collaborative ethical review, adopting internal as well as reciprocal monitoring schemes and encouraging task shifting within the teams.
  • WHO needs to update its good clinical practice guidelines, which were issued in 1995 (Lang et al. 2011) and consequently do not address most of the contemporary challenges faced by the noncommercial sponsors. The updated guidelines could, among others, allow new cost-effective quality control mechanisms such as internal monitoring, delegating tasks from the sponsor to the sites and double ethical review, to improve the protection of patients and communities.

Acknowledgements

We would like to thank: The consortium partners: the IRSS/Centre Muraz, Bobo Dioulasso, Burkina Faso; the Medical Research Unit, Hospital Albert Schweitzer, Lambarene, Gabon; the Barcelona Centre for International Health Research (CRESIB, Hospital Clínic- University of Barcelona), Spain; the Manhica Health Research Centre (CISM), Mozambique; the Institute for Tropical Diseases Research and Prevention of the University of Calabar Teaching Hospital, Nigeria; the National Malaria Control Programme, Kigali, Rwanda; the Uganda Malaria Surveillance Project, Kampala (UMSP), Uganda; Epicentre Paris, representing the affiliated centre in Mbarara, Uganda; the Tropical Diseases Research Centre, Ndola, Zambia; the Institute of Tropical Medicine of the University of Tubingen, Germany; the Liverpool School of Tropical Medicine, UK, and the University of Liverpool, UK. The staff at the 4-ABC research sites: Nanoro, Burkina Faso; Fougamou and Lambaréné, Gabon; Afokang and Pamol, Nigeria; Mashesha and Rukara, Rwanda; Jinja, Tororo and Mbarara, Uganda; Ndola, Zambia; and Manhiça, Mozambique. At Institute of Tropical Medicine, Antwerp, Belgium: Sharleen Braham, Christophe Burm, Yves Claeys, Jan Dierckx, Pascale Forret, Greta Gondol, Tom Koyen, Alessandro Massella, Evi Pockelé, Vera Van Boxel, Bjorn Van Den Sande, Danielle Van Melle, Jef Verellen, Peter Vermeulen. Thanks also to Prof Bruno Gryseels, director of the ITM, for his continuous support.

Raouf Osseni of Pharmaclin, Benin. The European Developing Countries Clinical Trials Partnership (EDCTP) which funded this study; the Medicine for Malaria Venture (MMV) that provided additional funding for the genotyping of blood samples; and the Belgian Directorate-General for Development Cooperation (DGD) that provided support through the framework agreement programme between DGD and ITMA. Sigma Tau for kindly providing DHAPQ and Sanofi Aventis for ASAQ. MR4 for providing us with malaria parasites which served as controls in the genotyping method in the laboratory part of the study.

Ancillary